How the census was built.

Two universes: active for headlines, active plus between-intakes for the index.

GrantCompass tracks 650+ Canadian business funding programs. The census draws two distinct universes from that catalog, and every figure states which one it uses.

All headline percentages and medians in the report are computed over active programs. The Open Door Index and the top-50 leaderboard add the 112 between-intakes programs — temporarily paused but expected to reopen — each badged "Between intakes." Closed and discontinued programs are excluded entirely. Where a field is not reported for every program, the report states the denominator (for example, "n=425 of 437 reporting"); we never impute a missing value to inflate coverage.

What the open data contains: aggregate tables and a top-50 leaderboard.

The open data is aggregate statistics only — distributions, shares and medians across the catalog, never a per-program export. Aggregating every named program is the GrantCompass product, so individual program rows are not published. Two things are visible: the downloadable aggregate tables (below), and a 50-program leaderboard rendered on the report page itself.

2a · Aggregate statistics tables

The downloadable files (census-2026-aggregates.json and census-2026-aggregates.csv) carry the tables below. The CSV is a single stacked ("tidy") file: each table is a block whose first column, table, names which table the row belongs to; a header row precedes each block. The JSON is the structured equivalent. Every table that aggregates an incompletely-covered field carries its denominator n.

A meta block records the generation date, license, attribution string, universe rules, and conventions (shares are fractions in 0–1; median = lower-middle value for even counts).

2b · The top-50 leaderboard fields

The report page shows "The 50 most open doors in Canada" — the 50 highest Open Door Index programs (active + between-intakes), as visible page content. Each leaderboard entry carries exactly these ten fields; no other per-program data is exposed.

Premium intelligence — exact competition scores, exact application hours, approval rates, realistic-amount estimates, eligible-expense detail, insider tips, application steps and the matching engine — is the paid product and is published in neither the aggregate tables nor the leaderboard.

Why we publish tiers, not exact ratings.

The Open Door Index inputs come from two derived buckets. We publish the bucket — not the underlying exact rating — because the exact competition score and exact application-hours estimate are part of the paid product. The buckets leak directional information only, while making the index reproducible.

Boundary convention: 9.9 hours falls in light, 10 and 40 hours in moderate, and 40.1 hours in heavy. A program whose hours are published only as a range (for example, "200–350") is treated as unrated for effort, not coerced to a number — so it scores blank rather than guessed.

A 0–100 measure of how accessible a program is.

The Open Door Index answers one question: how easy is it for a typical applicant to actually get this funding? It is computed only from publishable fields, and it uses fixed weights — not cross-program ranking — so any single program's score depends only on its own fields and reproduces in isolation.

A program scores only if all five inputs resolve: both derived tiers, both booleans, and a recognized payment model. A program missing any required input gets a blank index — no imputation — and is shown with a dash. Scores are rounded to one decimal place.

Computing the index for a fictional program.

Example — a fictional program (illustration only; not a real program's withheld values). Imagine a provincial grant with low competition, moderate effort, friendly to first-time applicants, no matching funds required, and paid by reimbursement — concretely: competition rated 2 of 5, an estimated 12 hours to apply.

A score of 76.3 places this program in the "accessible" band: low competition and no matching pull it up, while the reimbursement payment model and moderate effort keep it short of a perfect score. The single change that would lift it most is payment timing — if the same program paid an advance instead of reimbursing, paymentTiming would move from 0.25 to 1.0 and the index would rise to 87.5.

Hand-researched against official program pages.

Every program in the catalog is researched by hand against the funder's official page, with a verification date recorded per program. We do not scrape aggregator sites or accept second-hand summaries for the fields that drive the index. When a funder updates eligibility, amounts or status, the record is re-verified and the verification date advances. Programs that close or are discontinued are moved out of the active universe rather than left to drift.

Derived fields — competitionTier, effortTier and openDoorIndex — are computed deterministically from the verified source fields by the published build scripts. The same scripts produce both the statistics in the report and the downloadable aggregate tables, so the two cannot diverge. The index module ships with a regression test suite that locks the score for fixed inputs, so a code change that would alter any program's score fails the build.

CC BY 4.0 — free to use with attribution.

The Canadian Funding Census dataset and the figures in the report are released under Creative Commons Attribution 4.0 International. You may copy, redistribute, adapt and build on the data for any purpose, including commercially, provided you credit GrantCompass (grantcompass.ca) and indicate any changes you made.

Suggested attribution: "Data: GrantCompass Canadian Funding Census 2026 (grantcompass.ca), CC BY 4.0." Download the aggregate tables from the report's Data Hub. The published data is aggregate statistics only; the per-program product layer — exact competition and effort ratings, approval-odds detail, step-by-step playbooks, eligible-expense breakdowns, insider tips and the matching engine — is not part of the open data and is not licensed for reuse.

Version history.

The census is refreshed quarterly, or sooner when at least 20% of the catalog changes substantively. Each release is dated and recorded here so prior figures remain interpretable.

Ready to find funding that fits your business? Take the GrantCompass precision quiz and get a personalized list of programs ranked by your Open Door Index — filtered to your province, industry, and business stage.

← Back to the Canadian Funding Census 2026